The Obesity Paradox and Predictors of Thirty Day Mortality in a Single Center Heart Failure Population

Purpose Mortality associated with congestive heart failure continues to be a substantial problem. While in-hospital mortality has declined recently, this same trend has not been replicated in 30-day mortality, an important quality marker for many institutions. Thus, it is important to try to identify patient characteristics that would be predictive of increased mortality in order to target these risk factors to improve 30-day mortality. Methods We performed a retrospective chart review of 5885 patients admitted to Froedtert Hospital from January 2015 to August 2017. Patient data from these admissions was obtained from Froedtert's EMR. Patients were selected with a primary discharge diagnosis of acute decompensated heart failure, but were excluded if discharged to hospice. We then determined patient status at 30 days after discharge (alive or dead). We then analyzed several initial patient characteristics, including intake laboratory values and vital signs, current comorbidities, functional scores, and BMI. Each of these was analyzed utilizing a chi-square analysis to determine if they were independent predictors of 30-day mortality. Results After chi-square analysis, increased age and BUN, low systolic BP, active cancer, ejection fraction (EF) Conclusion In this study, we show that BMI, age, BUN, systolic BP, active cancer, EF, active smoking status and Walter score were independently predictive of 30-day mortality. Future research can look to utilize these characteristics as a score to predict 30-day mortality. We also found evidence of the obesity paradox in our population as well; At 30 days, patients with higher BMI had lower mortality rates, suggestive of a protective property of higher BMI. Further investigation is needed as to why.